The ontogram-approach to text processing and semantic relation spotting for indexing

This paper describes the OntoGram-approach to indexing texts by their conceptual content using ontologies along with syntactic grammars and lexico-syntactic information and semantic role assignment provided by lexical resources. The conceptual content of meaningful chunks of text is transformed into concept feature structures and mapped into concepts in a generative ontology. By this approach, synonymous but linguistically quite distinct expressions are mapped to the same concept in the ontology. This allows us to perform a content-based search which will retrieve relevant documents independently of the linguistic form of the query as well as the documents.